Government data can often reveal surprising insights about the way communities are served. In this project, built only using open data such as weather history, census data and New York’s 311 call data, we are trying to determine wether some parts of the city were served faster than others. Then we built a model to predict - at the time of the call - if the response to it will end up late or not.
The journey from the data collection to the predictive application will cover all the different steps of a data project such as data preparation or feature engineering. Several approaches to visualize the data will also be explored. A special focus will be on how to make different technologies (SQL, python, R) work together to get the best of all worlds.
Speaker: Maxime Mouchet